Deep learning-based classification of lumbar T2-weighted MRIs to subjects with and without low back pain symptoms

Deep learning-based classification of lumbar T2-weighted MRIs to subjects with and without low back pain symptoms | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Deep learning-based classification of lumbar T2-weighted MRIs to subjects with and without low back pain symptoms Mustafa Al-Rubaye, Satu Inkinen, Jaro Karppinen, Miika Nieminen, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6367177/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Low back pain (LBP) is a prevalent condition, with most individuals experiencing it at some point. Magnetic resonance imaging (MRI) is the primary imaging modality for LBP diagnostics, enabling visualization of bones, intervertebral discs, and neural structures. While MRI can reveal degenerative changes associated with LBP, these findings are also observed in asymptomatic individuals, limiting its diagnostic specificity. Purpose: This study investigates the feasibility of deep learning for automated classification of lumbar spine MRIs into symptomatic and asymptomatic cases. The proposed method could assist in LBP diagnostics by either detecting LBP presence or ruling out negative cases, streamlining the diagnostic workflow. Study Design: A deep learning-based classification approach using a pre-trained ResNet50 convolutional neural network (CNN) was developed to distinguish between symptomatic and asymptomatic cases. Methods: Sagittal T2-weighted MRIs from the Northern Finland Birth Cohort 1966 dataset were used to train, validate, and test the ResNet50-based model for LBP classification. The impact of varying the number of sagittal slices per subject was also evaluated. Results: The best classification performance was achieved using a modified pre-trained ResNet50 network. The model attained a Balanced Accuracy of \((0.713 \pm 0.014)\) , an Average Precision of \((0.494 \pm 0.013)\) , and an Area Under the Receiver Operating Characteristic Curve (AUC-ROC) of \((0.724 \pm 0.004)\) . Binary classification was more effective in predicting asymptomatic cases, while incorporating Pfirrmann grading improved symptomatic case predictions. Conclusions: Deep learning-based classification of lumbar spine MRIs can distinguish symptomatic and asymptomatic cases, with potential applications in LBP diagnostics. Deep Learning Image Classification Magnetic Resonance Imaging Lumbar Spine Low Back Pain Full Text Additional Declarations No competing interests reported. Supplementary file Appendix A is not available with this version. Supplementary Files OnlineGraphicalAbstractarticle.png The workflow of the classification process designed in the present study. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6367177","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":448885084,"identity":"6261299c-6049-41db-a282-fafed3101e04","order_by":0,"name":"Mustafa Al-Rubaye","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCklEQVRIiWNgGAWjYNACGyBmB2LGBhADSDxgYOBnw60eqC4NSDHDtPAcYGBIYGCQbCNei0QCREsDDvW6M9KfP/iRYMfA38zA9pl3h11iv+TjYxKJbQwSfDi0mN3IMWzsSUhmkDjMwDyb90xy4szZaWlgLbj8AtTC2MD7A+iqw/yfmXnbmBM33M4xA2mpw60l/WHjn4R6BnmgLUAt9Yn7b54xI2BLgmEzT8JhBgOIlsOJGyR4CGg588ZwtkzCcR5DoBbGuWeOG884k5ZskXBOAreW4+kPPr5JqJaTO97AzPB2R7Vsf/vhgzc+lNlIyDfg0AMFPDCGI1ShBH71yMCeeKWjYBSMglEwUgAAqypUKMTkjwsAAAAASUVORK5CYII=","orcid":"","institution":"University of Oulu","correspondingAuthor":true,"prefix":"","firstName":"Mustafa","middleName":"","lastName":"Al-Rubaye","suffix":""},{"id":448885085,"identity":"49aa88d7-0a87-46db-b686-2ad7771285d7","order_by":1,"name":"Satu Inkinen","email":"","orcid":"","institution":"Helsinki University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Satu","middleName":"","lastName":"Inkinen","suffix":""},{"id":448885086,"identity":"4f447844-239f-4b71-bd5a-591b9cc0d074","order_by":2,"name":"Jaro Karppinen","email":"","orcid":"","institution":"University of Oulu","correspondingAuthor":false,"prefix":"","firstName":"Jaro","middleName":"","lastName":"Karppinen","suffix":""},{"id":448885087,"identity":"a0955d4e-ec41-4356-a152-f65b12bca850","order_by":3,"name":"Miika Nieminen","email":"","orcid":"","institution":"University of Oulu","correspondingAuthor":false,"prefix":"","firstName":"Miika","middleName":"","lastName":"Nieminen","suffix":""},{"id":448885088,"identity":"0e74abbe-b59b-4509-9b11-2a1b0e051523","order_by":4,"name":"Juuso Ketola","email":"","orcid":"","institution":"Helsinki University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Juuso","middleName":"","lastName":"Ketola","suffix":""}],"badges":[],"createdAt":"2025-04-03 08:08:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6367177/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6367177/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100594331,"identity":"24cc91ca-42ad-47bb-b631-eb13353634e2","added_by":"auto","created_at":"2026-01-19 13:39:41","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8574239,"visible":true,"origin":"","legend":"","description":"","filename":"blindedSpringerNatureLaTeXTemplateLBP.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6367177/v1_covered_407907b3-de48-4a02-913a-b547833c5f3f.pdf"},{"id":82028960,"identity":"e94f4198-0aeb-42e8-8211-4ffc2cdb286c","added_by":"auto","created_at":"2025-05-06 07:08:11","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":8016402,"visible":true,"origin":"","legend":"The workflow of the classification process designed in the present study.","description":"","filename":"OnlineGraphicalAbstractarticle.png","url":"https://assets-eu.researchsquare.com/files/rs-6367177/v1/5b76721628ccbedbb27365a2.png"}],"financialInterests":"\u003cp\u003eNo competing interests reported.\u003c/p\u003e\n\u003cp\u003eSupplementary file Appendix A is not available with this version.\u003c/p\u003e","formattedTitle":"Deep learning-based classification of lumbar T2-weighted MRIs to subjects with and without low back pain symptoms","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Deep Learning, Image Classification, Magnetic Resonance Imaging, Lumbar Spine, Low Back Pain","lastPublishedDoi":"10.21203/rs.3.rs-6367177/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6367177/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e\u003cp\u003eLow back pain (LBP) is a prevalent condition, with most individuals experiencing it at some point. Magnetic resonance imaging (MRI) is the primary imaging modality for LBP diagnostics, enabling visualization of bones, intervertebral discs, and neural structures. While MRI can reveal degenerative changes associated with LBP, these findings are also observed in asymptomatic individuals, limiting its diagnostic specificity.\u003c/p\u003e\u003ch2\u003ePurpose:\u003c/h2\u003e\u003cp\u003eThis study investigates the feasibility of deep learning for automated classification of lumbar spine MRIs into symptomatic and asymptomatic cases. The proposed method could assist in LBP diagnostics by either detecting LBP presence or ruling out negative cases, streamlining the diagnostic workflow.\u003c/p\u003e\u003ch2\u003eStudy Design:\u003c/h2\u003e\u003cp\u003eA deep learning-based classification approach using a pre-trained ResNet50 convolutional neural network (CNN) was developed to distinguish between symptomatic and asymptomatic cases.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e\u003cp\u003eSagittal T2-weighted MRIs from the Northern Finland Birth Cohort 1966 dataset were used to train, validate, and test the ResNet50-based model for LBP classification. The impact of varying the number of sagittal slices per subject was also evaluated.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e\u003cp\u003eThe best classification performance was achieved using a modified pre-trained ResNet50 network. The model attained a Balanced Accuracy of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((0.713 \\pm 0.014)\\)\u003c/span\u003e\u003c/span\u003e, an Average Precision of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((0.494 \\pm 0.013)\\)\u003c/span\u003e\u003c/span\u003e, and an Area Under the Receiver Operating Characteristic Curve (AUC-ROC) of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\((0.724 \\pm 0.004)\\)\u003c/span\u003e\u003c/span\u003e. Binary classification was more effective in predicting asymptomatic cases, while incorporating Pfirrmann grading improved symptomatic case predictions.\u003c/p\u003e\u003ch2\u003eConclusions:\u003c/h2\u003e\u003cp\u003eDeep learning-based classification of lumbar spine MRIs can distinguish symptomatic and asymptomatic cases, with potential applications in LBP diagnostics.\u003c/p\u003e","manuscriptTitle":"Deep learning-based classification of lumbar T2-weighted MRIs to subjects with and without low back pain symptoms","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-06 07:08:05","doi":"10.21203/rs.3.rs-6367177/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0f1c4c07-5e4d-44d4-9887-82e0d362330c","owner":[],"postedDate":"May 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-17T08:09:18+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-06 07:08:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6367177","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6367177","identity":"rs-6367177","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}